Manage entities
Last updated
Last updated
You can manage entity types and entities on Knowledge. Knowledge data will be shared with your other Scopes.
The Literal Entity Type consists of multiple Literal Value Blocks. Literal Value used in API interaction and multiple entities that are possible inputs from your customers.
Literal Value Block: Each block consist of pairs of one Literal Value and single or multiple Entities.
Literal Value: Functions similarly to a key.
Entity: Corresponds to the value. The difference from the Regex Entity Type is that the latter uses regular expressions for the entity value, while the Literal Entity Type simply uses a string.
“Lemmatize” feature makes listing-up the entity value easier. When you don’t activate “lemmatize” feature, if you take the entity value “knife”, AI don’t identify “knives” with the entity value, “knife”. Because the literal form of “knife” and “knives” is not same. But when you activate it, “knife” is enough to cover “knife” and “knives”.
When a word is used in sentences, the word can be presented in various forms. For example, the verb to go can be present as goes in she goes to school every day and as went in she went to school yesterday, and as going in she is going to school now. But as we know, the basic form of them which is used as a dictionary headword is go. The basic form of a word like go is called lemma. And lemmatization means making the various forms into a lemma of the word.
Please refer the examples below.
Lemma | Variation | |
---|---|---|
Noun | candle | candle, candles |
Verb | go | go, goes, went, gone, going |
Adjective | clear | clear, clearer, clearest |
There's a "Case Sensitive" feature that can be turned on or off. When it's on, exact matches between an input and a predefined entity value are required for recognition. This exact match doesn't just concern the sequence of letters but also their specific case (uppercase or lowercase). So, for example, "apple" and "Apple" would be considered different when this feature is active.
Add Entity Types
Click Add Entity Type.
Select the composition as Literal.
Enter Entity Type name and description.
Click Case Sensitive checkbox.
Enter values for Literal Value Block.
Click Add.
Edit Entity Types
Click the view more menu of the Entity Type.
Click Edit
Click Case Sensitive checkbox.
Click Save.
Click Add Entity Type.
Select the composition as Literal.
Enter entity type name and description.
Enter values for Literal Value Block.
Click Add.
Open the view more menu of the entity type.
Click Edit.
Edit and click Save.
Open the view more menu of the entity type.
Click Delete.
Open the view more menu of the block.
Click Edit.
Edit and click Save.
Open the view more menu of the block.
Click Delete.
A regex entity type holds Regex Pattern Blocks which consist of a Regular Expression search pattern and its description.
Click Add Entity Type.
Select the composition as Regex.
Enter entity type name and description.
Enter Regex Pattern Block values.
Click Add.
Open the view more menu of the entity type.
Click Edit.
Edit and click Save.
Open the view more menu of the entity type.
Click Delete.
Open the view more menu of the block.
Click Edit.
Edit and click Save.
Open the view more menu of the block.
Click Delete.
You can view system entities that are eligible for extension in the System (Extensible) tab. You can extend the entity types by adding blocks or extending entities.
Choose an entity type to extend from the list.
Click Add Literal Value Block (for literal entity type) or Add Regex Pattern Block (for regex entity type).
Enter values for the block.
Click Add.
Extending entities is only eligible for literal entity types.
Choose an entity type to extend from the list.
Click Extend Entity from the block to extend.
Enter entities.
Click Extend.
Open the view more menu of the entity extension.
Click Edit.
Edit and click Save.